Considering that the "cliff" is defined based on CAPIE (i.e. CA area relative to IJIS extent), it's not only not surprising it doesn't show up on IJIS, it's a logical impossibility for it to do so!
Chris Reynolds thinks (and I agree) that the "cliff" is in large part due to melt ponding. This lowers area but not extent, for two reasons:
(a) melt ponds cause the measured concentration to drop to intermediate levels ~50% that affect area but don't trip the extent threshold of 15%
(b) CA and IJIS use different algorithms to turn microwave brightness into ice concentration measurement. CA uses the same algorithm year-round, while IJIS tunes its summer algorithm to be less affected by melt ponding.
Thus, it's not surprising that the "cliff" showed up in the last week or so, concurrent with the onset of melt ponding across the Arctic fringes - first in the CA, then in the Laptev/East Siberian Sea, and now in the Chukchi and Beaufort. Depending how abrupt the onset of melt ponding in the central basin is, there may well be more cliff to come.
More interesting is what was happening back at the start of the month, when there was a mini-cliff (remember: defined as area dropping off relative to extent), at a time when there was <i>not</i> any melt ponding in the Central Arctic - no darkening visible by MODIS, and all mass balance buoys still showing substantial snow. This it seems to me was the effect of the persistent cyclone in diverging and melting the ice, causing area to go down but not tripping the area threshold.